143
J. Hyg., Camb. (1979), 82, 143 Printed in Great Britain
Salmonellosis in Indonesia: phage type distribution of Salmonella typhi* BY WARREN R. SANBORN U.S. Naval Medical Research Unit No. 2, Jakarta Detachment, APO San Francisco, California 96356 J.-F. VIEU Centre Franpais de Lysotypie Enterique, Institut Pasteur, Paris
S. KOMALARINI AND SINTA Sumber Waras Hospital, Jakarta R. TRENGGONOWATI AND I. L. KADIRMAN Fatmawati Hospital, Cilandak, Jakarta SUMARMO AND ABDEL AZIZ Department of Gastroenterology, Medical School, University of Indonesia, Jakarta T. SADJIMIN AND TRIWIBOWO Faculty of Medicine, Gadja Mada University, Yogyakarta
ATAS AND SUTRISNA AYAM Public Health Laboratory, Yogyakarta
(Received 15 May 1978) SUMMARY
The distribution of phage types was studied among 577 strains of Salmonella typhi from Indonesia. Chemotype, colicinogeny, and tetrathionate reductase activity were also studied for most of these strains. The current phage type formula for Java was determined to be: A, D2, D6, Ela, E2, Ml, and 46, but two other large groups of strains were also found, I + IV and degraded Vi + strains. Significant differences in S. typhi strain distributions were noted between two localities on Java with respect to phage type and tetrathionate reductase activity. Comparisons were made with past phage typing studies in Jakarta as well as with more recent studies in other parts of south-east Asia. Phage types A, Dl, D2, and El persisted at a rather steady level in Jakarta for 28 years. Evidence was found for * The views presented in this paper are those of the authors. No endorsement by the Department of the Navy has been given or should be inferred. This project was supported by the U.S. Naval Medical Research and Development Command, Work Unit MR51. 524.009-0040. Reprint requests to Publications Office, NAMRU-2, Box 14, APO San Francisco 96263 or 7-1 Kung Yuan Road, Taipei, Taiwan, Republic of China.
0022-1724/79/0061-1978 $01.00 © 1979 Cambridge University Press
144
W. R. SANBORN AND OTHERS
epidemiological links to European and Asian areas. Antibiotic resistance among these Indonesian S. typhi strains was rare. INTRODUCTION Typhoid fever is a major public health problem of developing countries. It is also increasing elsewhere in places such as France and the United States, possibly associated with travel from Asia, the Middle East, Africa, and South America (Vieu, Diverneau & Binette, 1976; Rice, Baine & Gangarosa, 1977). Since Cragie & Yen (1938) described specific adaptation of certain bacteriophages to S. typhi strains of different origins, their phage-typing method has evolved into a major tool for epidemiological study of the sources of typhoid fever (Anderson & Williams, 1956). It is also an important method to demonstrate the spread of typhoid fever throughout the world (Felix, 1955; Nicolle, 1962a, b; Anderson & Nicolle, 1973). Erber & DeMoor (1954) observed that in the tropics, where typhoid fever is endemic, phage typing has special value to help trace specific outbreaks in epidemiological studies. The recent application of computer techniques has greatly facilitated such epidemiologic studies (Vieu et al. 1973). Distribution of S. typhi phage types in certain areas of Indonesia has been reported previously (Lie Kian Joe, 1949; Lie Kian Joe, 1950; Erber & DeMoor, 1954; Felix, 1955; Gan-Oei Siok Yauw & Lie Kian Joe, 1960). The present study was designed to update information regarding S. typhi phage type distribution on Java and to assess the variation within Java as well as over other parts of southeast Asia. MATERIALS AND METHODS This study was based on 577 strains of S. typhi isolated from human sources during the period September 1973 through July 1975. Most were from blood culture of acute cases of enteric fever, but two strains were isolated from stools of carriers in Jakarta. There were 232 strains from the Jakarta area, these coming from four hospitals drawing patients from the northern, western, and central (2) areas of Jakarta. An additional 108 strains came from a hospital in suburban Cilandak that received patients from the suburban-rural area on the south side of Jakarta. The remaining 237 strains were obtained from the Yogyakarta area in central Java. A modification of the standard method of phage-typing (Cragie & Yen, 1938; Cragie & Felix, 1947) was used as described by Nicolle (1961). There were 96 Vi phages used in this study as distributed by Dr E. S. Anderson from the Inter-
national Centre for Enteric Phage Typing (ICEPT). A complementary phage typing system was used to subdivide phage type A strains (Nicolle, Pavlatou & Diverneau, 1953). Several other complementary differentiating tests were done on these strains in addition to phage typing. All strains were examined for chemotype, i.e. acid production from xylose in peptone water with bromthymol blue, 24 h at 37 'C. Chemotype I was xylose positive while chemotype II was negative. The
Salmonellosis in Indonesia
145
colicinogeny of the S. typhi strains against Escherichia coli K-12 was also studied (Nicolle & Prunet, 1964). Additionally, most strains were tested for their ability to reduce tetrathionate (TTR test) (Nicolle & LeMinor, 1965). In collating the variety of data collected during this study, a computer data bank was formed from which phage type frequencies were determined (Vieu et al. 1973). The frequency of S. typhi phage types at various localities and for different times were compared using the x2 test, considering the data as a 2 x R table. Individual pair comparisons of specific phage types or TTR types were examined using a test for the significance of difference between two sample proportions. In this case a conservative acceptance level was adopted (P < 0.001), since the overall protection level (probability that any of the obtained differences were due to chance) for all possible tests made at the 0 001 significance level was approximately 0-02 for this study. Protection level was calculated as 1- (1-(c-) where a = significance level and K = number of possible comparisons (Ferguson, 1966). When using the hypothesis testing format to establish the fact that two groups of phage types of TTR types did not differ significantly, it was necessary to allow for a very liberal acceptance level 'a' (e.g. P > 0.25), thus reducing ,B error (probability that a real difference is not detected) to a reasonable amount. RESULTS There were 16 phage types found among the 577 strains of S. typhi in this study. Additionally, I+ IV, degraded Vi + and Vi negative strains were present. The largest single group, 25 %, consisted of strains designated I + IV, those resistant to Vi11 typing phages but reacting with I and IV unadapted Vi phages. The distribution of these 19 groups in Jakarta, Cilandak, and Yogyakarta appears in Table 1. A comparison of phage type distribution was made between Jakarta and its suburb, Cilandak, phage type 53 being omitted from this analysis, since it was the cause of a focal outbreak that was the subject of intense case finding efforts that might have biased the result. The distribution of phage types was not significantly different between these two areas (P > 0.30). Therefore, in all subsequent analyses, results from these areas were grouped together as the Jakarta area. Phage type distributions were then compared for two diverse geographic areas of Java, Jakarta at the western end and centrally located Yogyakarta (Table 2). The five predominant types were the same for each locality. These were I+ IV, D2, A, degraded Vi + strains and Ela in decreasing order of prevalence. Vi negative strains were sixth most common in Jakarta and seventh in Yogyakarta. In Jakarta 6 2 % of strains were phage type M1, while only 0 4 % of this type was found in Yogyakarta (P < 0.001). Phage type E2 was 6 8 % of Yogyakarta strains with only 0-4 % being found in Jakarta (P < 0 001). The remaining distributional differences were composed of small numbers of strains. During this study 11 strains of S. typhi were isolated that were resistant to chloramphenicol, streptomycin, sulpha drugs, and tetracycline (CSSuT). Ten of IO
HYG 82
W. R. SANBORN AND
146
OTHERS
Table 1. Phage type distribution of Salmonella typhi strains isolated from human sources on Java, 1973-5 Phage type/group
Jakarta 64
I+IV Degraded Vi + Vi negative A C2 DI D2 D6 Ela Elb E2 GI G3
Cilandak 32 15
18 23 25 1 3 40 9 20 2
48 20 9 40
4 12
1
12 4 9 1
54 15 18 16 2
1
1
2 14
Ji Ml
25 42 46 53 Strains examined *
Yogyakarta*
1974-75.
7
1
4 2
1
10
7
lot
232
237
108
t A focal outbreak.
Table 2. Comparative distribution of Salmonella typhi phage types from diverse geographic areas on Java, 1974-5 Phage type/group I+IV Degraded Vi+ Vi negative A C2 DI D2 D6 Ela Elb E2 Gl G3 Ji Ml
25 42 46 53
R = < 0-5%.
Jakarta area, 253 strains (%) 32 11 6*3 8'7 R 0*8 14 4'0 91 1-2 R
Yogyakarta, 237 strains
(%) 20 8*4 3-8 17 R 23 6-3 7-6 6*8 0'8
R 0'8 5.1
3*2 4.0* * Focal outbreak. Blank
R 1*7 0*8 3-0 space,
type not found.
147
Salmonellosis in Indonesia Table 3. Tetrathionate reductase (TTR) tests on 394 strains of Salmonella typhi collected from humans 1973-5 TTR negative
Phage type I+IV Degraded Vi+ Vi negative D2 D6 E2 53 Totals
(±) 2 13 (1) 3 1 (1) 2 1 10 31(2)
Strains tested 144 53 36 106 28 17 10 394
these were from a focal outbreak in Cilandak (Sanborn et al. 1975) and were phage type 53, the only members of this phage type found. The other was from an isolated case in central Jakarta and was a I + IV strain. Sub-types of phage type A found in this study were Tananarive-25, Chamblee-7, Dakar-2, and non-typable strains-3 for the Jakarta area while there were 31, 5, 1, and 3 respectively for the Yogyakarta area. Thus, there were no appreciable distributional differences. There were 509 S. typhi strains examined for chemotype. The majority, 403, were chemotype I, xylose positive. Ratios of chemotypes I and II from the three areas, Jakarta, Cilandak, and Yogyakarta, were not significantly different. Chemotype I predominated among I + IV, degraded Vi +, and Vi negative strains as well as among all phage types except for C2, D6, Ml, and 42. Among strains of these phage types the chemotype I to II ratios were 0: 1, 1: 3, 1: 10, and 0: 2 respectively. Again, there were no significant differences between the three geographic areas. There were 394 S. typhi strains tested for tetrathionate reductase activity, and 31 were negative with two plus-minus reactions. These were distributed among phage types as shown in Table 3. Jakarta yielded 5 TTR negative strains, Cilandak 23 (1 +) and Yogyakarta 3 (1 +). Of the Cilandak strains, 10 were from the focal outbreak of phage type 53. The proportions of TTR negative strains in Jakarta and Cilandak were not significantly different, whether including the phage type 53 strains (P > 006) or excluding these (P > 0.44). However, comparing the Jakarta-Cilandak distribution to that of Yogyakarta revealed a highly significant difference (P < 0.001). Colicinogeny was tested on 572 strains, and all were negative. The present Jakarta phage type distribution as compared with that of earlier studies is shown in Table 4 (Felix, 1955; Gan-Oei Siok Yauw & Lie Kian Joe, 1960; Lie Kian Joe, 1949). Phage types A, Dl, D2, El, and the rare E2 were found in all these studies. The same was true for degraded Vi + and Vi negative strains. Phage type M (Ml) was common in our study at 6-2 % where it had been about 2-4 % in the previous studies (P < 0.001). A significant decrease was seen in the amount of phage type D6, dropping from a major position at approximately 24 to 28 % over three previous studies down to 3-8 % in the present study (P < 0.001). Phage type BI was not found in Jakarta, where it had been as frequent as 8 % of IO-2
W. R. SANBORN AND OTHERS
148
Table 4. Historical comparison of Salmonella typhi phage type distribution in Jakarta, Indonesia 1973-5
Phage types Rank Jakarta 1973-5 1 I+IV 2 D2 3 A 4 Degraded Vi + 5 El (Ela) 6 Vi negative 7 M (MI) 8 D6 9 46 10 53 11 Dl 12 J (Ji) 13 C2 14 E2 15 G (G3) Other phage types found, 1947-60
% 28-2 15-3 10*4 9-7
9.4 7.9 6-2 3-8 3-2 2-9
0*9 0-6 R R R
Non-typable E3 25 E4 Bi D10 D9 36 D7 29 40 N Strains examined
340
1955-60*
1947-8t
1950-2t Rank
%
22'5 14-0 6-8 5*3 3*6 0-7 24-4
3 2 9 4 10 8 1
13-4 14-5 3.5 10-7 2-1 3-6 28-0
2 3
21-4 20-4
4 8 10 1
12-6 1-8 0.7 27-4
0.5
11
1P3
9
1P4
Rank
%
2 3 5 6 8 11 1 14
§
Rank
§
%
§
R 10
1-4
12
1.0
11
4 7 9 9 12 13 15 16 17 18 19
8*7 4-8 2'4 1-7 07 0*7 05 05 R R R
5
8-6
7
3-2
7
4.7
6
3-9
6
8-0
5
6-7
12
R
414
910
R
285
Blank space, not found; -, phage not available; R, < 0.5 %. * Adapted from Gan-Oei Siok Yauw & Lie Kian Joe (1960). t Adapted from Felix (1955). 1 Adapted from Lie Kian Joe (1949). § Designation not in use.
strains in previous reports. Phage type 25 also was not found in Jakarta, but a few of these strains were found in Yogyakarta. Current distributions of S. typhi phage types were compared for Java and the Indochina peninsula (Nicolle et al. 1975). The results are shown in Table 5. Major strain types common to the two areas were I + IV, degraded Vi positive, and El. Analysis of the distributions of the following phage types showed highly significant differences (P < 0 001) between the two areas: D2, D6, 46, E2, A, Vi negative, Ml, and GI. Some phage types (EIO, N, 37, T, 29, B2, and E9), common on the Indochina peninsula, were not found in this study.
Salmonellosis in Indonesia
149
Table 5. Comparative distribution of Salmonella typhi phage types in southeast Asian areas Java 1973-5 Phage types Java 1973-5 I+IV D2 A Degrades Vi + El Vi negative D6 Ml 46 E2 53 25 DI J1 GI 42 G3 C2
Other phage types found in south east Asia EIO N 37 T 29 B2 E9 J3 M2 E3 BI Cl
Philippinesl A
Indochina Peninsula*t 1948-73
1969-72
1954-9
Rank
%
Rank
%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
25*0 18'4 13*3 9*2 8'7 6*2 4.9 3*8 3*1 2*9 1-7 0*7
1 6 3 2 5 8 16 4 29 35
28*3 4-6 8-7
4 1
0*8 68-3
10*8 7*0 3.4
5
0*8
2 3 6
2
23*6
7 4
4.3
28 18 14 9
R R 1-5
6
0*8
44 30
R R
7 10 11 12 13 15 17 19 20 21 26 34
3-8 2*7 2*4 2*0 1*5 1*2 1*2
7
0*8
4.9
1
550
3
0*6 R R R R R
1*2 8-7 R R
Rank
%
Rank 5
%
2.1(?) 33.7 4.3 2-1
1.1
§
2*8
0*9 0*9 0*8 R R
124 4514 577 Strains examined 282 Blank space, Phage type not found; R, < 0 5%; (?), 'Non-typable', presumed I+ IV. * Cambodia, Laos, Thailand, Vietnam. t Adapted from Nicolle et al. (1975). $ Adapted from Basaca-Sevilla et at. (1974). § Not found in this collection but reported by others from Thailand.
Distribution of phage types from Java was compared with that of the Philippines (Basaca-Sevilla et al. 1974). Type BI was most frequent in the Philippines but not found in Java, and type A was much more prevalent in the Philippines than in Indonesia (Table 5).
150
W. R. SANBORN AND OTHERS DISCUSSION
The 577 strains reported here comprised a sample comparable to previous studies, only that by Felix (1955) being substantially larger. Numbers of strains collected from Jakarta and Yogyakarta were generally proportionate to the populations in those areas. Jakarta strains provided a good cross-section of the community, coming from five hospitals serving four areas of the city. Similarly, Yogyakarta strains came from the major hospital there as well as from clinics in suburban areas surrounding the city. The present basic phage type formula for Java appears to be A, D2, D6, Eta, E2, M1, and 46, plus I + IV and degraded Vi + strains. This pattern is similar to that reported by Felix (1955) and also by Lie Kian Joe (1949), except for the lack of types Bi, E3, and E4; the lower frequency of Dl and 25; and the addition of the newer type 46. Thus, phage types A, Dl, D2, D6, El, E2, and M have maintained endemic presence in Jakarta for at least 28 years. I + IV strains formed the largest single group in this study. They were proportionately similar to the distribution seen on the Indochina peninsula (Nicolle et al. 1975). Before 1955 this designation was not in use, but even the 'non-typable' category in earlier studies could not account for so large a group, since nontypables were only one-third the amount of I+ IV strains found here. Thus, it would seem that the increase in these strains represented a real change in southeast Asian S. typhi distribution. This study documented the occurrence of the newer phage types 46 and 53 in Indonesia for the first time. These types were not available for use in the earlier studies, having been distributed by the International Reference Laboratory for Enteric Phage Typing in 1961 and 1972 respectively. Elsewhere, phage type 46 has been reported primarily from Europe, with the exception of France. Its highest numbers were generally found in England, the Netherlands, Belgium, Spain, and central Europe (Anderson & Nicolle, 1973). Previous studies indicated that it had been rare in south-east Asia (Nicolle et al. 1975), but in the present study it ranked ninth at 31 %. It may be that this prevalence could be due to Indonesia's close ties to the Netherlands through business, travel, and trade. The recently recognized phage type 53 was a unique finding. It made its only appearance in this study as a confined outbreak of 10 typhoid cases. The source of this outbreak remained obscure despite our intense epidemiological studies. Phage type 53 was introduced into the international system of phage typing by ICEPT in 1972. Therefore, its true frequency and distribution is difficult to define. Following its discovery in Japan, phage type 53 was reported from diverse places including Sweden and Thailand (Anderson, 1974; Anderson, 1976). It has been found by the Pasteur Institute among recent S. typhi collections from South Korea, Sudan, Central African Republic, Iraq, and France (J.-F. Vieu, personal communication). It appears that phage type 53 may be rather widely distributed. Thus, this study suggests two possible typhoid contaminations of Indonesia from external sources, phage types 46 and 53. Phage type 25 was originally discovered in Indonesia (Lie Kian Joe, 1949), and
Salmonellosis in Indonesia
151
in subsequent studies it continued to represent a considerable position of the S. typhi strains isolated in Jakarta. However, since no strains of this type were found in Jakarta and only a few in Yogyakarta, it seems that this phage type is becoming less frequent in Java. Another significant decline was noted in the frequency of phage type D6. It was the predominant type in the three previous studies, representing 24, 28, and 27 % of the strains isolated. In this study it ranked eighth at 4 %. Finding the cosmopolitan phage type Ela was not unexpected. It has been common in previous Jakarta studies and throughout Indochina. However, chemotype analysis revealed that one-third of our strains were mainly from the Jakarta area. Sub-typing the phage type A strains revealed a preponderance of Tananarive among the three sub-types and untypable strains found. There were also three strains of Dakar, a sub-type usually restricted to West Africa. There was a significant difference between the phage type distributions of Jakarta and Yogyakarta. Phage type E2 was much more prevalent in Yogyakarta, 6.8 % v. 0.4 %, and M1 was more prevalent in Jakarta, 6 2 % v. 0 4 %. Phage type MI has been relatively common in the Far East and is generally found around the borders of the Pacific Ocean, including Japan (Anderson & Nicolle, 1973). Its occurrence in Jakarta was similar to, but somewhat lower than, that found on the Indochina peninsula (Nicolle et al. 1975), perhaps reflecting partial contact between this major centre of Indonesian commerce and other parts of Asia. The fact that it was rare in Yogyakarta provided evidence that central Java might be somewhat epidemiologically isolated. The frequency of phage type E2 in Yogyakarta was even more interesting. This phage type has been limited to Cambodia, Viet Nam, and the Philippines. In the Philippines it was isolated frequently from 1953 through 1958 and again in 1972 (Basaca-Sevilla et at. 1974). It is possible that an epidemiological link exists between the Philippines and Yogyakarta, one not shared bv Jakarta and the Philippines. Strains positive in the colicinogeny test are very rare, except for phage type 40, where all strains are colicinogenic (Nicolle & Prunet, 1964). This phage type was not found in Indonesia; our strains were all negative, as expected. Apart from the rare chemotype of phage type Ela mentioned earlier, the two chemotypes were evenly distributed throughout the culture collection. Shifts in this pattern, an easily performed test, could prove to be a useful epidemiological reference. Of the 394 strains tested by the TTR test (Nicolle & LeMinor, 1965) 0-8 % were negative. These strains were examined for distribution among phage types as well as geographical distribution. As with the phage type distribution, there was no significant difference in the TTR negative strain distribution between Jakarta and Cilandak. However, significantly fewer TTR negative S. typhi strains were found in Yogyakarta. This finding supports other evidence for the epidemiological or ecological isolation of Yogyakarta, as noted above and with S. paratyphi A strains (Sanborn et al. 1977).
152
W. R. SANBORN AND OTHERS
In conclusion, the epidemiology of S. typhi on Java shows several features: (1) A current phage type formula: A, D2, D6, El, E2, Ml, and 46, plus I+IV and degraded Vi + strains. (2) Persistence over 28 years of the following phage types: A, Dl, D2, D6, El, E2, M (Ml). (3) I + IV strains predominant. (4) Significant differences in phage type and tetrathionate reductase-negative strain distributions between geographically separated localities on the same island, Java. (5) Reduction in phage type 25 strains, a type originally discovered in Indonesia. (6) Antibiotic resistance rare among these Indonesian S. typhi strains. We gratefully acknowledge Capt. P. F. D. Van Peenen, M.C., U.S.N., who provided a group of S. typhi strains collected in late 1973 and also Dr Guno Wiseso who arranged for specimens from the Infectious Diseases Hospital in Jakarta. The capable technical assistance of Mme G. Diverneau and Mlle H. Binette in characterizing strains is recognized with gratitude. We also acknowledge the careful laboratory studies by Anni Chiang, Windani Chandra, Sotyo Iswara, Mulyadis, and Sanhawi. Statistical advice was given by Dr C. W. Hutchins, and computer services were made available by the Naval Health Research Center,
San Diego, California. REFERENCES ANDERSON, E. S. (1974). Salmonella surveillance. Chloramphenicol-resistant Salmonella typhi in Viet-Nam (Rep. of) and Thailand. Weekly Epidemiological Record 41, 295. ANDERSON, E. S. (1976). Salmonella phage types and drug resistance in 1975. Weekly Epidemiological Record 51, 317. ANDERSON, E. S. & NIcoLLE, P. (1973). Geographical distribution of Salmonella typhi and S. fparatyphi A and B phage types during the period 1 January 1966 to 31 December 1969. A report of the International Committee for Enteric Phage Typing. Journal of Hygiene 71, 59. ANDERSON, E. S. & WILLIAMS, R. E. 0. (1956). Bacteriophage typing of enteric pathogens and staphylococci and its use in epidemiology: A review. Journal of Clinical Pathology 9, 94. BASACA-SEVILLA, V., GAETOS, D., MANIAGO, E., SEVILLA, J. & ZACARIAS, L. (1974). Salmonella and Shigella serotypes from human cases and contacts from 1969 to 1972 in the Philippines. Philippine Society Medicine and Infectious Di8eaae Journal 3, 38. C1AGiE, J. & FELIX, A. (1947). Typing of typhoid bacilli with Vi bacteriophage. Lancet i, 823. CRAGIE, J. & YEN, C. H. (1938). Demonstration of types of B. typhosus by means of preparations of type II Vi phage: principles and technique. Canadian Public Health Journal 29, 448. ERBER, M. & DEMooR, C. E. (1954). Typhoid fever and phage typing on the east coast of Sumatra. Documenta de Medicina Geographica et Tropica 6, 144. FELIX, A. (1955). World survey of typhoid and paratyphoid phage types. Bulletin of the World Health Organization 13, 109. FERGUSON, G. A. (1966). Statistical Analysis in P8ychology and Education, 2nd ed. New York: McGraw-Hill Book Co. GAN-OEI SIOK YAIUW & LIE KIAN JOE (1960). Hasil Penentuan type phaag daripada Salmonella typhi de Djakarta. Madgalah Kedokteran Indonesia 10, 453. LIE KLiN JOE (1 949). Typing of typhoid bacilli from Indonesia by means of Vi-Bacteriophage II. Documenta Neerlandica et Indonesica de Morbis Tropici8 1, 145. Ln3 KiAN JOE (1950). Further results obtained in typing typhoid bacilli in Indonesia and the West Indies by the phage-method. Documenta Neerlandica et Indonesica de Morbis Tropicia 2, 176.
Salmonellosis in Indonesia
153
NICOTTR, P. (1961). La lysotypie de Salmonella typhi, son principe, sa technique, son application a l'epidemiologie de la fievre typhoide, 67. In E. van Oye (ed.), The World Problem of Salmonellos8i, vitgeverij Dr W. Junk, den Haag. NicoTTi, P. (1962 a, b). Rapport sur la distribution des lysotypes de S. typhi et de S. paratyphi B dans le monde d'apres les resultats fournis par les centres nationaux membres du Comit6 International de Lysotypie Ent6rique a l'occasion du Congres International de Microbiologies, Stockholm, 1958. Annales de l'Institut Pasteur, Paris 102, (a) 389; (b) 580. NIcoTT , P. & LEMINOR, L. (1965). Sur la pr6sence ou l'absence de la reductase du tetrathionate dans une collection de bacille typhique de provenance variee. Int6ret 6pid6miologique 6ventuel de la recherche de cette enzyme. Annales de l'Institut Pasteur, Paris 108, 501. NicoLLE, P., PAVLATOU, M. & DIVERNEAU, G. (1953). Subdivision de quelques types Vi frequents de Salmonella typhi par des lysotypies auxiliaires. Comptes Rendus Academie des Sciences 236, 2453. NICOLLE, P. & PRUNET, J. (1964). La propri6t6 colicinogene dans l'espece S. typhi. Annales de l'Institut Pasteur, Paris 107, 174. NICOLLE, P., VIEu, J.-F., SOK SoN, K. H., DIvERNEAU, G. & BRAULT, J. (1975). Recherche de nouveaux lysotypes de Salmonella typhi parmi les nombreuses souches Vi positives non lysotypables (groupe I + IV) provenand de la p6ninsule indochinoise. Revue Epidemiologie, Medicine Sociale et Sante Publique 23, 193. RIcE, P. A., BAINE, W. B. & GANGAROSA, E. J. (1977). Salmonella typhi infections in the United States, 1967-1972: Increasing importance of international travelers. American Journal of Epidemiology 106, 160. SANBORN, W. R., HABLAS, R., KoMATIANI, S., SINTA, TRENGGONOWATI, R., SADJIMIN, T., ATAS & SUTRISNA (1977). Salnonellosis in Indonesia: phage type distribution of Salmonella paratyphi A. Journal of Hygiene 79, 1. SANBORN, W. R., LESMANA, M., DENNIS, D. T., TRENGGONOWATI, R., KEnIDRmAN, I. L. & SuANI (1975). Antibiotic-resistant typhoid in Indonesia. Lancet ii, 408. VIEu, J.-F., DivERNEAU, G. & BiNsa'IE, H. (1976). La fi6vre typhoide en France: Lysotypie de 529 souches de Salmonella typhi isol6es en 1975. Medicine et Maladies Infectienses 6, 260. VIEu, J.-F., TOUCAS, M., FRECHIN, C. & SCIOT, A. (1973). Application de l'ordinateur a l'6pid6miologie de Salmonella typhi, Salmonella paratyphi B et Salmonella typhi-murium en France: Elaboration d'un thesaurus 6pid6miologique. Revue Epidemiologie, Medicine Sociale et Sante Publique 21, 51.
J. Hyg., Camb. (1979), 82, 143 Printed in Great Britain
Salmonellosis in Indonesia: phage type distribution of Salmonella typhi* BY WARREN R. SANBORN U.S. Naval Medical Research Unit No. 2, Jakarta Detachment, APO San Francisco, California 96356 J.-F. VIEU Centre Franpais de Lysotypie Enterique, Institut Pasteur, Paris
S. KOMALARINI AND SINTA Sumber Waras Hospital, Jakarta R. TRENGGONOWATI AND I. L. KADIRMAN Fatmawati Hospital, Cilandak, Jakarta SUMARMO AND ABDEL AZIZ Department of Gastroenterology, Medical School, University of Indonesia, Jakarta T. SADJIMIN AND TRIWIBOWO Faculty of Medicine, Gadja Mada University, Yogyakarta
ATAS AND SUTRISNA AYAM Public Health Laboratory, Yogyakarta
(Received 15 May 1978) SUMMARY
The distribution of phage types was studied among 577 strains of Salmonella typhi from Indonesia. Chemotype, colicinogeny, and tetrathionate reductase activity were also studied for most of these strains. The current phage type formula for Java was determined to be: A, D2, D6, Ela, E2, Ml, and 46, but two other large groups of strains were also found, I + IV and degraded Vi + strains. Significant differences in S. typhi strain distributions were noted between two localities on Java with respect to phage type and tetrathionate reductase activity. Comparisons were made with past phage typing studies in Jakarta as well as with more recent studies in other parts of south-east Asia. Phage types A, Dl, D2, and El persisted at a rather steady level in Jakarta for 28 years. Evidence was found for * The views presented in this paper are those of the authors. No endorsement by the Department of the Navy has been given or should be inferred. This project was supported by the U.S. Naval Medical Research and Development Command, Work Unit MR51. 524.009-0040. Reprint requests to Publications Office, NAMRU-2, Box 14, APO San Francisco 96263 or 7-1 Kung Yuan Road, Taipei, Taiwan, Republic of China.
0022-1724/79/0061-1978 $01.00 © 1979 Cambridge University Press
144
W. R. SANBORN AND OTHERS
epidemiological links to European and Asian areas. Antibiotic resistance among these Indonesian S. typhi strains was rare. INTRODUCTION Typhoid fever is a major public health problem of developing countries. It is also increasing elsewhere in places such as France and the United States, possibly associated with travel from Asia, the Middle East, Africa, and South America (Vieu, Diverneau & Binette, 1976; Rice, Baine & Gangarosa, 1977). Since Cragie & Yen (1938) described specific adaptation of certain bacteriophages to S. typhi strains of different origins, their phage-typing method has evolved into a major tool for epidemiological study of the sources of typhoid fever (Anderson & Williams, 1956). It is also an important method to demonstrate the spread of typhoid fever throughout the world (Felix, 1955; Nicolle, 1962a, b; Anderson & Nicolle, 1973). Erber & DeMoor (1954) observed that in the tropics, where typhoid fever is endemic, phage typing has special value to help trace specific outbreaks in epidemiological studies. The recent application of computer techniques has greatly facilitated such epidemiologic studies (Vieu et al. 1973). Distribution of S. typhi phage types in certain areas of Indonesia has been reported previously (Lie Kian Joe, 1949; Lie Kian Joe, 1950; Erber & DeMoor, 1954; Felix, 1955; Gan-Oei Siok Yauw & Lie Kian Joe, 1960). The present study was designed to update information regarding S. typhi phage type distribution on Java and to assess the variation within Java as well as over other parts of southeast Asia. MATERIALS AND METHODS This study was based on 577 strains of S. typhi isolated from human sources during the period September 1973 through July 1975. Most were from blood culture of acute cases of enteric fever, but two strains were isolated from stools of carriers in Jakarta. There were 232 strains from the Jakarta area, these coming from four hospitals drawing patients from the northern, western, and central (2) areas of Jakarta. An additional 108 strains came from a hospital in suburban Cilandak that received patients from the suburban-rural area on the south side of Jakarta. The remaining 237 strains were obtained from the Yogyakarta area in central Java. A modification of the standard method of phage-typing (Cragie & Yen, 1938; Cragie & Felix, 1947) was used as described by Nicolle (1961). There were 96 Vi phages used in this study as distributed by Dr E. S. Anderson from the Inter-
national Centre for Enteric Phage Typing (ICEPT). A complementary phage typing system was used to subdivide phage type A strains (Nicolle, Pavlatou & Diverneau, 1953). Several other complementary differentiating tests were done on these strains in addition to phage typing. All strains were examined for chemotype, i.e. acid production from xylose in peptone water with bromthymol blue, 24 h at 37 'C. Chemotype I was xylose positive while chemotype II was negative. The
Salmonellosis in Indonesia
145
colicinogeny of the S. typhi strains against Escherichia coli K-12 was also studied (Nicolle & Prunet, 1964). Additionally, most strains were tested for their ability to reduce tetrathionate (TTR test) (Nicolle & LeMinor, 1965). In collating the variety of data collected during this study, a computer data bank was formed from which phage type frequencies were determined (Vieu et al. 1973). The frequency of S. typhi phage types at various localities and for different times were compared using the x2 test, considering the data as a 2 x R table. Individual pair comparisons of specific phage types or TTR types were examined using a test for the significance of difference between two sample proportions. In this case a conservative acceptance level was adopted (P < 0.001), since the overall protection level (probability that any of the obtained differences were due to chance) for all possible tests made at the 0 001 significance level was approximately 0-02 for this study. Protection level was calculated as 1- (1-(c-) where a = significance level and K = number of possible comparisons (Ferguson, 1966). When using the hypothesis testing format to establish the fact that two groups of phage types of TTR types did not differ significantly, it was necessary to allow for a very liberal acceptance level 'a' (e.g. P > 0.25), thus reducing ,B error (probability that a real difference is not detected) to a reasonable amount. RESULTS There were 16 phage types found among the 577 strains of S. typhi in this study. Additionally, I+ IV, degraded Vi + and Vi negative strains were present. The largest single group, 25 %, consisted of strains designated I + IV, those resistant to Vi11 typing phages but reacting with I and IV unadapted Vi phages. The distribution of these 19 groups in Jakarta, Cilandak, and Yogyakarta appears in Table 1. A comparison of phage type distribution was made between Jakarta and its suburb, Cilandak, phage type 53 being omitted from this analysis, since it was the cause of a focal outbreak that was the subject of intense case finding efforts that might have biased the result. The distribution of phage types was not significantly different between these two areas (P > 0.30). Therefore, in all subsequent analyses, results from these areas were grouped together as the Jakarta area. Phage type distributions were then compared for two diverse geographic areas of Java, Jakarta at the western end and centrally located Yogyakarta (Table 2). The five predominant types were the same for each locality. These were I+ IV, D2, A, degraded Vi + strains and Ela in decreasing order of prevalence. Vi negative strains were sixth most common in Jakarta and seventh in Yogyakarta. In Jakarta 6 2 % of strains were phage type M1, while only 0 4 % of this type was found in Yogyakarta (P < 0.001). Phage type E2 was 6 8 % of Yogyakarta strains with only 0-4 % being found in Jakarta (P < 0 001). The remaining distributional differences were composed of small numbers of strains. During this study 11 strains of S. typhi were isolated that were resistant to chloramphenicol, streptomycin, sulpha drugs, and tetracycline (CSSuT). Ten of IO
HYG 82
W. R. SANBORN AND
146
OTHERS
Table 1. Phage type distribution of Salmonella typhi strains isolated from human sources on Java, 1973-5 Phage type/group
Jakarta 64
I+IV Degraded Vi + Vi negative A C2 DI D2 D6 Ela Elb E2 GI G3
Cilandak 32 15
18 23 25 1 3 40 9 20 2
48 20 9 40
4 12
1
12 4 9 1
54 15 18 16 2
1
1
2 14
Ji Ml
25 42 46 53 Strains examined *
Yogyakarta*
1974-75.
7
1
4 2
1
10
7
lot
232
237
108
t A focal outbreak.
Table 2. Comparative distribution of Salmonella typhi phage types from diverse geographic areas on Java, 1974-5 Phage type/group I+IV Degraded Vi+ Vi negative A C2 DI D2 D6 Ela Elb E2 Gl G3 Ji Ml
25 42 46 53
R = < 0-5%.
Jakarta area, 253 strains (%) 32 11 6*3 8'7 R 0*8 14 4'0 91 1-2 R
Yogyakarta, 237 strains
(%) 20 8*4 3-8 17 R 23 6-3 7-6 6*8 0'8
R 0'8 5.1
3*2 4.0* * Focal outbreak. Blank
R 1*7 0*8 3-0 space,
type not found.
147
Salmonellosis in Indonesia Table 3. Tetrathionate reductase (TTR) tests on 394 strains of Salmonella typhi collected from humans 1973-5 TTR negative
Phage type I+IV Degraded Vi+ Vi negative D2 D6 E2 53 Totals
(±) 2 13 (1) 3 1 (1) 2 1 10 31(2)
Strains tested 144 53 36 106 28 17 10 394
these were from a focal outbreak in Cilandak (Sanborn et al. 1975) and were phage type 53, the only members of this phage type found. The other was from an isolated case in central Jakarta and was a I + IV strain. Sub-types of phage type A found in this study were Tananarive-25, Chamblee-7, Dakar-2, and non-typable strains-3 for the Jakarta area while there were 31, 5, 1, and 3 respectively for the Yogyakarta area. Thus, there were no appreciable distributional differences. There were 509 S. typhi strains examined for chemotype. The majority, 403, were chemotype I, xylose positive. Ratios of chemotypes I and II from the three areas, Jakarta, Cilandak, and Yogyakarta, were not significantly different. Chemotype I predominated among I + IV, degraded Vi +, and Vi negative strains as well as among all phage types except for C2, D6, Ml, and 42. Among strains of these phage types the chemotype I to II ratios were 0: 1, 1: 3, 1: 10, and 0: 2 respectively. Again, there were no significant differences between the three geographic areas. There were 394 S. typhi strains tested for tetrathionate reductase activity, and 31 were negative with two plus-minus reactions. These were distributed among phage types as shown in Table 3. Jakarta yielded 5 TTR negative strains, Cilandak 23 (1 +) and Yogyakarta 3 (1 +). Of the Cilandak strains, 10 were from the focal outbreak of phage type 53. The proportions of TTR negative strains in Jakarta and Cilandak were not significantly different, whether including the phage type 53 strains (P > 006) or excluding these (P > 0.44). However, comparing the Jakarta-Cilandak distribution to that of Yogyakarta revealed a highly significant difference (P < 0.001). Colicinogeny was tested on 572 strains, and all were negative. The present Jakarta phage type distribution as compared with that of earlier studies is shown in Table 4 (Felix, 1955; Gan-Oei Siok Yauw & Lie Kian Joe, 1960; Lie Kian Joe, 1949). Phage types A, Dl, D2, El, and the rare E2 were found in all these studies. The same was true for degraded Vi + and Vi negative strains. Phage type M (Ml) was common in our study at 6-2 % where it had been about 2-4 % in the previous studies (P < 0.001). A significant decrease was seen in the amount of phage type D6, dropping from a major position at approximately 24 to 28 % over three previous studies down to 3-8 % in the present study (P < 0.001). Phage type BI was not found in Jakarta, where it had been as frequent as 8 % of IO-2
W. R. SANBORN AND OTHERS
148
Table 4. Historical comparison of Salmonella typhi phage type distribution in Jakarta, Indonesia 1973-5
Phage types Rank Jakarta 1973-5 1 I+IV 2 D2 3 A 4 Degraded Vi + 5 El (Ela) 6 Vi negative 7 M (MI) 8 D6 9 46 10 53 11 Dl 12 J (Ji) 13 C2 14 E2 15 G (G3) Other phage types found, 1947-60
% 28-2 15-3 10*4 9-7
9.4 7.9 6-2 3-8 3-2 2-9
0*9 0-6 R R R
Non-typable E3 25 E4 Bi D10 D9 36 D7 29 40 N Strains examined
340
1955-60*
1947-8t
1950-2t Rank
%
22'5 14-0 6-8 5*3 3*6 0-7 24-4
3 2 9 4 10 8 1
13-4 14-5 3.5 10-7 2-1 3-6 28-0
2 3
21-4 20-4
4 8 10 1
12-6 1-8 0.7 27-4
0.5
11
1P3
9
1P4
Rank
%
2 3 5 6 8 11 1 14
§
Rank
§
%
§
R 10
1-4
12
1.0
11
4 7 9 9 12 13 15 16 17 18 19
8*7 4-8 2'4 1-7 07 0*7 05 05 R R R
5
8-6
7
3-2
7
4.7
6
3-9
6
8-0
5
6-7
12
R
414
910
R
285
Blank space, not found; -, phage not available; R, < 0.5 %. * Adapted from Gan-Oei Siok Yauw & Lie Kian Joe (1960). t Adapted from Felix (1955). 1 Adapted from Lie Kian Joe (1949). § Designation not in use.
strains in previous reports. Phage type 25 also was not found in Jakarta, but a few of these strains were found in Yogyakarta. Current distributions of S. typhi phage types were compared for Java and the Indochina peninsula (Nicolle et al. 1975). The results are shown in Table 5. Major strain types common to the two areas were I + IV, degraded Vi positive, and El. Analysis of the distributions of the following phage types showed highly significant differences (P < 0 001) between the two areas: D2, D6, 46, E2, A, Vi negative, Ml, and GI. Some phage types (EIO, N, 37, T, 29, B2, and E9), common on the Indochina peninsula, were not found in this study.
Salmonellosis in Indonesia
149
Table 5. Comparative distribution of Salmonella typhi phage types in southeast Asian areas Java 1973-5 Phage types Java 1973-5 I+IV D2 A Degrades Vi + El Vi negative D6 Ml 46 E2 53 25 DI J1 GI 42 G3 C2
Other phage types found in south east Asia EIO N 37 T 29 B2 E9 J3 M2 E3 BI Cl
Philippinesl A
Indochina Peninsula*t 1948-73
1969-72
1954-9
Rank
%
Rank
%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
25*0 18'4 13*3 9*2 8'7 6*2 4.9 3*8 3*1 2*9 1-7 0*7
1 6 3 2 5 8 16 4 29 35
28*3 4-6 8-7
4 1
0*8 68-3
10*8 7*0 3.4
5
0*8
2 3 6
2
23*6
7 4
4.3
28 18 14 9
R R 1-5
6
0*8
44 30
R R
7 10 11 12 13 15 17 19 20 21 26 34
3-8 2*7 2*4 2*0 1*5 1*2 1*2
7
0*8
4.9
1
550
3
0*6 R R R R R
1*2 8-7 R R
Rank
%
Rank 5
%
2.1(?) 33.7 4.3 2-1
1.1
§
2*8
0*9 0*9 0*8 R R
124 4514 577 Strains examined 282 Blank space, Phage type not found; R, < 0 5%; (?), 'Non-typable', presumed I+ IV. * Cambodia, Laos, Thailand, Vietnam. t Adapted from Nicolle et al. (1975). $ Adapted from Basaca-Sevilla et at. (1974). § Not found in this collection but reported by others from Thailand.
Distribution of phage types from Java was compared with that of the Philippines (Basaca-Sevilla et al. 1974). Type BI was most frequent in the Philippines but not found in Java, and type A was much more prevalent in the Philippines than in Indonesia (Table 5).
150
W. R. SANBORN AND OTHERS DISCUSSION
The 577 strains reported here comprised a sample comparable to previous studies, only that by Felix (1955) being substantially larger. Numbers of strains collected from Jakarta and Yogyakarta were generally proportionate to the populations in those areas. Jakarta strains provided a good cross-section of the community, coming from five hospitals serving four areas of the city. Similarly, Yogyakarta strains came from the major hospital there as well as from clinics in suburban areas surrounding the city. The present basic phage type formula for Java appears to be A, D2, D6, Eta, E2, M1, and 46, plus I + IV and degraded Vi + strains. This pattern is similar to that reported by Felix (1955) and also by Lie Kian Joe (1949), except for the lack of types Bi, E3, and E4; the lower frequency of Dl and 25; and the addition of the newer type 46. Thus, phage types A, Dl, D2, D6, El, E2, and M have maintained endemic presence in Jakarta for at least 28 years. I + IV strains formed the largest single group in this study. They were proportionately similar to the distribution seen on the Indochina peninsula (Nicolle et al. 1975). Before 1955 this designation was not in use, but even the 'non-typable' category in earlier studies could not account for so large a group, since nontypables were only one-third the amount of I+ IV strains found here. Thus, it would seem that the increase in these strains represented a real change in southeast Asian S. typhi distribution. This study documented the occurrence of the newer phage types 46 and 53 in Indonesia for the first time. These types were not available for use in the earlier studies, having been distributed by the International Reference Laboratory for Enteric Phage Typing in 1961 and 1972 respectively. Elsewhere, phage type 46 has been reported primarily from Europe, with the exception of France. Its highest numbers were generally found in England, the Netherlands, Belgium, Spain, and central Europe (Anderson & Nicolle, 1973). Previous studies indicated that it had been rare in south-east Asia (Nicolle et al. 1975), but in the present study it ranked ninth at 31 %. It may be that this prevalence could be due to Indonesia's close ties to the Netherlands through business, travel, and trade. The recently recognized phage type 53 was a unique finding. It made its only appearance in this study as a confined outbreak of 10 typhoid cases. The source of this outbreak remained obscure despite our intense epidemiological studies. Phage type 53 was introduced into the international system of phage typing by ICEPT in 1972. Therefore, its true frequency and distribution is difficult to define. Following its discovery in Japan, phage type 53 was reported from diverse places including Sweden and Thailand (Anderson, 1974; Anderson, 1976). It has been found by the Pasteur Institute among recent S. typhi collections from South Korea, Sudan, Central African Republic, Iraq, and France (J.-F. Vieu, personal communication). It appears that phage type 53 may be rather widely distributed. Thus, this study suggests two possible typhoid contaminations of Indonesia from external sources, phage types 46 and 53. Phage type 25 was originally discovered in Indonesia (Lie Kian Joe, 1949), and
Salmonellosis in Indonesia
151
in subsequent studies it continued to represent a considerable position of the S. typhi strains isolated in Jakarta. However, since no strains of this type were found in Jakarta and only a few in Yogyakarta, it seems that this phage type is becoming less frequent in Java. Another significant decline was noted in the frequency of phage type D6. It was the predominant type in the three previous studies, representing 24, 28, and 27 % of the strains isolated. In this study it ranked eighth at 4 %. Finding the cosmopolitan phage type Ela was not unexpected. It has been common in previous Jakarta studies and throughout Indochina. However, chemotype analysis revealed that one-third of our strains were mainly from the Jakarta area. Sub-typing the phage type A strains revealed a preponderance of Tananarive among the three sub-types and untypable strains found. There were also three strains of Dakar, a sub-type usually restricted to West Africa. There was a significant difference between the phage type distributions of Jakarta and Yogyakarta. Phage type E2 was much more prevalent in Yogyakarta, 6.8 % v. 0.4 %, and M1 was more prevalent in Jakarta, 6 2 % v. 0 4 %. Phage type MI has been relatively common in the Far East and is generally found around the borders of the Pacific Ocean, including Japan (Anderson & Nicolle, 1973). Its occurrence in Jakarta was similar to, but somewhat lower than, that found on the Indochina peninsula (Nicolle et al. 1975), perhaps reflecting partial contact between this major centre of Indonesian commerce and other parts of Asia. The fact that it was rare in Yogyakarta provided evidence that central Java might be somewhat epidemiologically isolated. The frequency of phage type E2 in Yogyakarta was even more interesting. This phage type has been limited to Cambodia, Viet Nam, and the Philippines. In the Philippines it was isolated frequently from 1953 through 1958 and again in 1972 (Basaca-Sevilla et at. 1974). It is possible that an epidemiological link exists between the Philippines and Yogyakarta, one not shared bv Jakarta and the Philippines. Strains positive in the colicinogeny test are very rare, except for phage type 40, where all strains are colicinogenic (Nicolle & Prunet, 1964). This phage type was not found in Indonesia; our strains were all negative, as expected. Apart from the rare chemotype of phage type Ela mentioned earlier, the two chemotypes were evenly distributed throughout the culture collection. Shifts in this pattern, an easily performed test, could prove to be a useful epidemiological reference. Of the 394 strains tested by the TTR test (Nicolle & LeMinor, 1965) 0-8 % were negative. These strains were examined for distribution among phage types as well as geographical distribution. As with the phage type distribution, there was no significant difference in the TTR negative strain distribution between Jakarta and Cilandak. However, significantly fewer TTR negative S. typhi strains were found in Yogyakarta. This finding supports other evidence for the epidemiological or ecological isolation of Yogyakarta, as noted above and with S. paratyphi A strains (Sanborn et al. 1977).
152
W. R. SANBORN AND OTHERS
In conclusion, the epidemiology of S. typhi on Java shows several features: (1) A current phage type formula: A, D2, D6, El, E2, Ml, and 46, plus I+IV and degraded Vi + strains. (2) Persistence over 28 years of the following phage types: A, Dl, D2, D6, El, E2, M (Ml). (3) I + IV strains predominant. (4) Significant differences in phage type and tetrathionate reductase-negative strain distributions between geographically separated localities on the same island, Java. (5) Reduction in phage type 25 strains, a type originally discovered in Indonesia. (6) Antibiotic resistance rare among these Indonesian S. typhi strains. We gratefully acknowledge Capt. P. F. D. Van Peenen, M.C., U.S.N., who provided a group of S. typhi strains collected in late 1973 and also Dr Guno Wiseso who arranged for specimens from the Infectious Diseases Hospital in Jakarta. The capable technical assistance of Mme G. Diverneau and Mlle H. Binette in characterizing strains is recognized with gratitude. We also acknowledge the careful laboratory studies by Anni Chiang, Windani Chandra, Sotyo Iswara, Mulyadis, and Sanhawi. Statistical advice was given by Dr C. W. Hutchins, and computer services were made available by the Naval Health Research Center,
San Diego, California. REFERENCES ANDERSON, E. S. (1974). Salmonella surveillance. Chloramphenicol-resistant Salmonella typhi in Viet-Nam (Rep. of) and Thailand. Weekly Epidemiological Record 41, 295. ANDERSON, E. S. (1976). Salmonella phage types and drug resistance in 1975. Weekly Epidemiological Record 51, 317. ANDERSON, E. S. & NIcoLLE, P. (1973). Geographical distribution of Salmonella typhi and S. fparatyphi A and B phage types during the period 1 January 1966 to 31 December 1969. A report of the International Committee for Enteric Phage Typing. Journal of Hygiene 71, 59. ANDERSON, E. S. & WILLIAMS, R. E. 0. (1956). Bacteriophage typing of enteric pathogens and staphylococci and its use in epidemiology: A review. Journal of Clinical Pathology 9, 94. BASACA-SEVILLA, V., GAETOS, D., MANIAGO, E., SEVILLA, J. & ZACARIAS, L. (1974). Salmonella and Shigella serotypes from human cases and contacts from 1969 to 1972 in the Philippines. Philippine Society Medicine and Infectious Di8eaae Journal 3, 38. C1AGiE, J. & FELIX, A. (1947). Typing of typhoid bacilli with Vi bacteriophage. Lancet i, 823. CRAGIE, J. & YEN, C. H. (1938). Demonstration of types of B. typhosus by means of preparations of type II Vi phage: principles and technique. Canadian Public Health Journal 29, 448. ERBER, M. & DEMooR, C. E. (1954). Typhoid fever and phage typing on the east coast of Sumatra. Documenta de Medicina Geographica et Tropica 6, 144. FELIX, A. (1955). World survey of typhoid and paratyphoid phage types. Bulletin of the World Health Organization 13, 109. FERGUSON, G. A. (1966). Statistical Analysis in P8ychology and Education, 2nd ed. New York: McGraw-Hill Book Co. GAN-OEI SIOK YAIUW & LIE KIAN JOE (1960). Hasil Penentuan type phaag daripada Salmonella typhi de Djakarta. Madgalah Kedokteran Indonesia 10, 453. LIE KLiN JOE (1 949). Typing of typhoid bacilli from Indonesia by means of Vi-Bacteriophage II. Documenta Neerlandica et Indonesica de Morbis Tropici8 1, 145. Ln3 KiAN JOE (1950). Further results obtained in typing typhoid bacilli in Indonesia and the West Indies by the phage-method. Documenta Neerlandica et Indonesica de Morbis Tropicia 2, 176.
Salmonellosis in Indonesia
153
NICOTTR, P. (1961). La lysotypie de Salmonella typhi, son principe, sa technique, son application a l'epidemiologie de la fievre typhoide, 67. In E. van Oye (ed.), The World Problem of Salmonellos8i, vitgeverij Dr W. Junk, den Haag. NicoTTi, P. (1962 a, b). Rapport sur la distribution des lysotypes de S. typhi et de S. paratyphi B dans le monde d'apres les resultats fournis par les centres nationaux membres du Comit6 International de Lysotypie Ent6rique a l'occasion du Congres International de Microbiologies, Stockholm, 1958. Annales de l'Institut Pasteur, Paris 102, (a) 389; (b) 580. NIcoTT , P. & LEMINOR, L. (1965). Sur la pr6sence ou l'absence de la reductase du tetrathionate dans une collection de bacille typhique de provenance variee. Int6ret 6pid6miologique 6ventuel de la recherche de cette enzyme. Annales de l'Institut Pasteur, Paris 108, 501. NicoLLE, P., PAVLATOU, M. & DIVERNEAU, G. (1953). Subdivision de quelques types Vi frequents de Salmonella typhi par des lysotypies auxiliaires. Comptes Rendus Academie des Sciences 236, 2453. NICOLLE, P. & PRUNET, J. (1964). La propri6t6 colicinogene dans l'espece S. typhi. Annales de l'Institut Pasteur, Paris 107, 174. NICOLLE, P., VIEu, J.-F., SOK SoN, K. H., DIvERNEAU, G. & BRAULT, J. (1975). Recherche de nouveaux lysotypes de Salmonella typhi parmi les nombreuses souches Vi positives non lysotypables (groupe I + IV) provenand de la p6ninsule indochinoise. Revue Epidemiologie, Medicine Sociale et Sante Publique 23, 193. RIcE, P. A., BAINE, W. B. & GANGAROSA, E. J. (1977). Salmonella typhi infections in the United States, 1967-1972: Increasing importance of international travelers. American Journal of Epidemiology 106, 160. SANBORN, W. R., HABLAS, R., KoMATIANI, S., SINTA, TRENGGONOWATI, R., SADJIMIN, T., ATAS & SUTRISNA (1977). Salnonellosis in Indonesia: phage type distribution of Salmonella paratyphi A. Journal of Hygiene 79, 1. SANBORN, W. R., LESMANA, M., DENNIS, D. T., TRENGGONOWATI, R., KEnIDRmAN, I. L. & SuANI (1975). Antibiotic-resistant typhoid in Indonesia. Lancet ii, 408. VIEu, J.-F., DivERNEAU, G. & BiNsa'IE, H. (1976). La fi6vre typhoide en France: Lysotypie de 529 souches de Salmonella typhi isol6es en 1975. Medicine et Maladies Infectienses 6, 260. VIEu, J.-F., TOUCAS, M., FRECHIN, C. & SCIOT, A. (1973). Application de l'ordinateur a l'6pid6miologie de Salmonella typhi, Salmonella paratyphi B et Salmonella typhi-murium en France: Elaboration d'un thesaurus 6pid6miologique. Revue Epidemiologie, Medicine Sociale et Sante Publique 21, 51.
